Towards a national primary health care data collection – dementia demonstration project

The project collected aggregate core general practice data items on dementia across 18 measures including comorbidities, risk factors and population demographics (for a full list of items, see Appendix A). This was achieved through the involvement, support and collaboration of participating Primary Health Networks (PHNs).

The project demonstrated that general practice data has the potential to contribute to an understanding of health conditions particularly where, as with dementia, there is limited information available from other sources.

The project also identified limitations in the coverage, availability, and quality of the data, and identified opportunities to improve the consistency, quality and accessibility of data. These improvements include developing and using consistent data standards and governance for the whole data lifecycle – from the point of collection to extraction and end use.

It also highlighted the importance of collaboration, the complexity of the primary health care data environment, and the challenges with developing a nationally consistent data collection.

These challenges included:

• comparability issues across data extractors and between data extracts
• lack of key data items available and extracted from clinical information systems
• lack of complete data item coverage across all PHNs
• variation in governance arrangements (including data suppression requirements) across PHNs
• lack of a standard data sharing agreement (DSA)
• variation between PHNs capacity to participate in the project, including resourcing.

The project used aggregate data, which limited the potential to study the underlying data quality. Using de-identified patient data in a future demonstration project would provide opportunity to assess the quality of the data collected in general practice systems.

The project used Primary Health Insights (PHI) to share data between PHNs and the Australian Institute of Health and Welfare (AIHW). PHI is a storage and analytics platform that hosts de-identified general practice and other primary health data of most PHNs. At the time of project implementation, some PHNs were still integrating PHI into existing workflows, developing their understanding of PHI and growing their technical capacity. Until this process is complete, this restricts the number of PHNs who are feasibly able to participate in projects that use PHI as the primary mechanism for data provision or collaborative analytics.

Dementia – a national priority in Australia

In 2023, an estimated 411,100 Australians were living with dementia (AIHW 2024). Ageing is the main risk factor, with the risk of developing dementia increasing for people aged 65 and over. It is estimated that 1.5% of Australians aged 30 and over and 8.4% aged 65 and over are living with dementia. An estimated 63% of Australians living with dementia are women (AIHW 2024).

Dementia was the second leading cause of death in 2022 (AIHW 2024) but there is a lack of national data on the diagnosis and management of dementia, particularly by primary care providers such as general practitioners (GPs), and specialists such as geriatricians.

Project findings

This project used 2 age groups to explore dementia in general practice:

• 65 years and over
• 30 years and over – analyses were limited to people aged 30 and over as dementia is a condition that primarily affects older people. This age limit is consistent with the lower age limit of dementia prevalence estimates for Australia (AIHW 2024).

The data showed that 0.8% of active general practice patients aged 30 and over, and 2.5% of those aged 65 and over, had a dementia diagnosis in 2023. 'Active' patients are those who had attended the general practice or service 3 or more times in the past 2 years.

The majority of participating PHNs reported that over 60% of general practices in their catchment area contributed data to the project. While all 17 participant PHNs provided data on patient demographics, due to privacy restrictions and/or quality issues, data could only be reported for 16 PHNs.

These findings should be interpreted with care and within the context of the project. They demonstrate the availability and utility of general practice data. However, the project presents preliminary findings only. The rates of dementia found by this project are not comparable with other estimates of people living with dementia because:

• The data for this project do not represent the broader population and were limited to:
  • 17 out of 31 PHNs
  • a subset of all general practices within participating PHNs
  • patients who had attended the general practice or service 3 or more times in the past 2 years
  • a practitioner either making a diagnosis of dementia for their patient, or being aware of a diagnosis from another source, and entering that diagnosis in their clinical information system in a way that it was included in the data extraction methodology for the system.
• There is no single authoritative data source for deriving dementia prevalence in Australia and different approaches are used to generate estimates. The prevalence estimates generated for the AIHW's Dementia in Australia report, for example, used a combination of data from the 2015 systematic review of worldwide dementia prevalence conducted by Alzheimer's Disease International for the World Alzheimer report 2015 (Prince et al. 2015), and an Australian study, The prevalence and causes of younger onset dementia in Eastern Sydney, Australia (Withall et al. 2014).

Although the rates of dementia from this project are not comparable with other estimates, it found similar demographic characteristics and comorbidities:

• A higher proportion of people with a dementia diagnosis were women. Women comprised a higher proportion than men in all 16 reportable PHNs (ranging from 52% to 61%).
• Dementia diagnoses increase with age. The dementia diagnosis rate in active patients aged 30 and over in the project was 0.8%, rising to 2.5% in active patients aged 65 and over.
• Comorbidities for people with dementia were similar to those found in other work based on other data sources. For example, chronic obstructive pulmonary disease (COPD) rates were between 7.7% and 11% for 13 of 16 PHNs in the project data set. Coronary heart disease rates ranged from 17% to 21% for 15 PHNs, and diabetes rates ranged from 18% to 34% across 16 PHNs. AIHW analysis of Australian Bureau of Statistics (ABS) Survey of Disability, Ageing and Carers 2018 data and ABS 2017–18 National Health Survey data similarly identified heart disease, diabetes and COPD as significant comorbidities in people with dementia (AIHW 2023, AIHW 2024).
• Dementia types identified in this study were consistent with those of other sources. Unspecified dementia, Alzheimer’s disease and vascular dementia were the most prevalent dementia types recorded for 10 PHNs. AIHW analysis of 2022–23 hospitalisations due to dementia found the most common dementia types included Alzheimer’s disease (30%), unspecified dementia (18%) and vascular dementia (13%) (AIHW 2024).

The project successfully identified which general practice data were available to understand dementia, and how these data could provide insight into the health of Australians. With further development, a general practice data set could contribute to an understanding of a range of other specific health conditions in the population.